Our UR5e screen calls a missed request a rejected bracket

CallumChen1143 · 1 Nov 2025, 14:58 UTC

Closed
CA
CallumChen1143
I watched the PLC send a short inspection request while the application was busy, then the training screen reported rejected when its wait expired; nobody can show me that the application accepted the request, and the ordinary job-interface clearing rules are still disputed (safety controls are separate).

15 replies

CA
CallumBrooks0795
Replying to CallumChen1143

Separate request acceptance from the inspection result. A timeout waiting for acceptance says nothing about the bracket's quality. Then define who owns and clears each part of the exchange.

25 points
ME
MeiAdams0168
Replying to CallumChen1143

Take rejected off that timeout path in the lesson draft now. We have been withholding a similar exercise until the normal and interrupted exchanges are agreed. It is a bad habit to teach confidently.

2 points
CA
CallumChen1143
Replying to CallumBrooks0795

The rejected wording is out of the draft; the application owner confirms it is only a timeout label, not an inspection result.

12 points
IS
IsabelAdams0126
Replying to CallumChen1143

Our reset-button review found the same problem from the other direction: a message action changed job state. Keep the request visible when explaining this timeout, so dismissing the message does not imply the work was cancelled.

9 points
HA
HazelChan1093
Replying to CallumChen1143

Could the operator tell whether to wait or call somebody? Unknown is more honest, but a screen full of honest words can still leave them guessing what to do next.

10 points
CA
CallumBrooks0795
Replying to IsabelAdams0126

Isabel's point matters if you add cancellation. A cancellation request needs an agreed response; do not treat pressing cancel as proof the receiver never accepted the inspection.

8 points
ME
MeiAdams0168
Replying to HazelChan1093

Hazel, yes. We used a paper screen with the person covering breaks and found the missing action immediately. Cheaper than discovering it during the practical session.

18 points
IS
IsabelAdams0126
Replying to MeiAdams0168

Mei, did your paper example use only information the real screen could obtain? I would not want a beautifully clear explanation of an acceptance state the interface never supplies.

5 points
ME
MeiAdams0168
Replying to IsabelAdams0126

That was a gap we had to fix. The walkthrough must use the actual agreed fields, not give the operator facts only the developers know. Thanks for making the limitation explicit.

7 points
CA
CallumChen1143
Replying to MeiAdams0168

Our current screen has no acceptance identity, just a connection indicator; I've asked both owners to define a matching acceptance and its retained information before we draw the recovery page.

6 points
HA
HazelChan1093
Replying to CallumChen1143

Then the connected indicator cannot answer the question. I would show the failed request wait plainly and route it to the named support role until the new exchange exists.

18 points
CA
CallumBrooks0795
Replying to CallumChen1143

Also agree whether the receiver durably retains an accepted request before the PLC clears it. An acknowledgement that disappears with the application process can leave the sender believing work is owned when it is not.

9 points
CA
CallumChen1143
Replying to CallumBrooks0795

That storage boundary is still open; the two owners have agreed one writer for each signal, but no revised exchange has been tested.

13 points
IS
IsabelAdams0126
Replying to CallumChen1143

Keep result unavailable distinct from no request accepted in the design too. They may both need attention, but the original bracket question differs.

12 points
HA
HazelChan1093
Replying to CallumChen1143

Will someone outside the design meeting try the next paper version? Let them tell you what they think happened before explaining it to them.

10 points

Discussion closed

This discussion is closed to new replies after six months without activity. Last activity: .